A Motion Planning Method for an Autonomous Underwater Vehicle with Kinematic Constraints

Motion planning is essential for the autonomous underwater vehicle (AUV) to achieve complete autonomous navigation in unknown environments. The two-layer motion planning framework, which combines path searching and trajectory parameterization, cannot deal with the non-zero initial state of AUV and is even unable to plan a dynamically feasible trajectory in an environment that needs reversing. In this paper, we combine the AUV motion state with the grid map and propose a path searching algorithm based on kinematics to ensure the correctness and dynamic feasibility of the path. By exploiting the advantages of the B-spline, the trajectory optimization strategy with a safety guarantee algorithm is proposed to ensure the safety of the final trajectory. Comparing with the other methods, the general applicability of this framework was demonstrated using a simulation system.